International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 36 - Number 8 |
Year of Publication: 2011 |
Authors: M.Ayman Al Ahmar |
10.5120/4511-6379 |
M.Ayman Al Ahmar . An Attribute-Rule Dependency Matrix Method and its Java Implementation for Rule-Based Expert Systems Verification. International Journal of Computer Applications. 36, 8 ( December 2011), 17-23. DOI=10.5120/4511-6379
Verification of knowledge bases is an important aspect of the development procedure of rule-based expert systems. The objective of verification is to assure producing a successful intelligent computer system that reaches correct recommendations. This research introduces an attribute-rule dependency matrix verification method and its associated Java implementation program. The method can help knowledge engineers and domain experts in the automated verification process of rule-based knowledge bases for both consistency and completeness. The method can also help in the documentation of expert systems' facts and If-Then rules. A wide variety of knowledge bases has been successfully debugged and analyzed using the introduced verification method.